Evaluating SAM2's Role in Camouflaged Object Detection: From SAM to SAM2

TL;DR

This paper evaluates SAM2's performance in camouflaged object detection, highlighting its improvements in speed and domain coverage over SAM, but also noting a decline in auto mode object perception without prompts.

Contribution

It introduces an assessment of SAM2's capabilities in camouflaged object detection, revealing strengths and limitations compared to SAM.

Findings

SAM2 outperforms SAM in speed and domain coverage.

SAM2's auto mode shows decreased object perception in camouflaged scenarios.

Results are available at the provided GitHub link.

Abstract

The Segment Anything Model (SAM), introduced by Meta AI Research as a generic object segmentation model, quickly garnered widespread attention and significantly influenced the academic community. To extend its application to video, Meta further develops Segment Anything Model 2 (SAM2), a unified model capable of both video and image segmentation. SAM2 shows notable improvements over its predecessor in terms of applicable domains, promptable segmentation accuracy, and running speed. However, this report reveals a decline in SAM2's ability to perceive different objects in images without prompts in its auto mode, compared to SAM. Specifically, we employ the challenging task of camouflaged object detection to assess this performance decrease, hoping to inspire further exploration of the SAM model family by researchers. The results of this paper are provided in \url{https://github.com/luckybird1994/SAMCOD}.